A report can contain valid findings while an engine, build or crawl stage remains incomplete. Presenting that result as either a total failure or a clean pass destroys useful information.
QFinch separates terminal status, stage status, coverage and findings. A warning tells the reviewer which planned work did not complete. Benchmark summaries preserve misses, false-positive candidates and unapproved truth mappings. Release decisions should require the relevant stages and evidence for that policy, rather than interpreting absence as proof.
This approach does not make a scanner accurate by declaration. It makes quality gaps visible enough to measure and improve.
Next step: Require coverage evidence beside zero-finding results in your release review.